[Feat] Detector implementation for Azure API Management Direct Management Key (#3938)

* azuredirectmanagementapikey detector implementation and tests

* add reference url

* use sanehttpclient

* better regex

* add prefix regex for key pattern

* some refactoring

* cleanup some unnecessary logic for expiry formatting.
code refactoring, remove unnecessary code.
updated the description of detector.
reduce the expiry to 5 seconds.

* improvise the regex to increase confidence level.
refactor variable name for better understanding.

* remove \b from regex

* implemented CustomFalsePositiveChecker interface.
implemented host caching if host is not reachable to avoid unnecessary extra verification calls.

* add delimiter to raw v2

* make protos

pkg/detectors/azuredirectmanagementkey/azuredirectmanagementkey.go

@@ -0,0 +1,163 @@
+package azuredirectmanagementkey
+
+import (
+	"context"
+	"crypto/hmac"
+	"crypto/sha512"
+	"encoding/base64"
+	"errors"
+	"fmt"
+	"net/http"
+	"strings"
+	"time"
+
+	regexp "github.com/wasilibs/go-re2"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/cache/simple"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
+	logContext "github.com/trufflesecurity/trufflehog/v3/pkg/context"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
+)
+
+const RFC3339WithoutMicroseconds = "2006-01-02T15:04:05"
+
+type Scanner struct {
+	client *http.Client
+	detectors.DefaultMultiPartCredentialProvider
+}
+
+// Ensure the Scanner satisfies the interface at compile time.
+var _ detectors.Detector = (*Scanner)(nil)
+var _ detectors.CustomFalsePositiveChecker = (*Scanner)(nil)
+
+var (
+	defaultClient = common.SaneHttpClient()
+	urlPat        = regexp.MustCompile(`https://([a-z0-9][a-z0-9-]{0,48}[a-z0-9])\.management\.azure-api\.net`)                                        // https://azure.github.io/PSRule.Rules.Azure/en/rules/Azure.APIM.Name/
+	keyPat        = regexp.MustCompile(detectors.PrefixRegex([]string{"azure", ".management.azure-api.net"}) + `([a-zA-Z0-9+\/]{83,85}[a-zA-Z0-9]==)`) // pattern for both Primary and secondary key
+
+	invalidHosts  = simple.NewCache[struct{}]()
+	noSuchHostErr = errors.New("no such host")
+)
+
+// Keywords are used for efficiently pre-filtering chunks.
+// Use identifiers in the secret preferably, or the provider name.
+func (s Scanner) Keywords() []string {
+	return []string{".management.azure-api.net"}
+}
+
+// FromData will find and optionally verify Azure Management API keys in a given set of bytes.
+func (s Scanner) FromData(ctx context.Context, verify bool, data []byte) (results []detectors.Result, err error) {
+	logger := logContext.AddLogger(ctx).Logger().WithName("azuredirectmanagementkey")
+	dataStr := string(data)
+
+	urlMatchesUnique := make(map[string]string)
+	for _, urlMatch := range urlPat.FindAllStringSubmatch(dataStr, -1) {
+		urlMatchesUnique[urlMatch[0]] = urlMatch[1] // urlMatch[0] is the full url, urlMatch[1] is the service name
+	}
+	keyMatchesUnique := make(map[string]struct{})
+	for _, keyMatch := range keyPat.FindAllStringSubmatch(dataStr, -1) {
+		keyMatchesUnique[strings.TrimSpace(keyMatch[1])] = struct{}{}
+	}
+
+EndpointLoop:
+	for baseUrl, serviceName := range urlMatchesUnique {
+		for key := range keyMatchesUnique {
+			s1 := detectors.Result{
+				DetectorType: detectorspb.DetectorType_AzureDirectManagementKey,
+				Raw:          []byte(baseUrl),
+				RawV2:        []byte(baseUrl + ":" + key),
+			}
+
+			if verify {
+				if invalidHosts.Exists(baseUrl) {
+					logger.V(3).Info("Skipping invalid registry", "baseUrl", baseUrl)
+					continue EndpointLoop
+				}
+
+				client := s.client
+				if client == nil {
+					client = defaultClient
+				}
+
+				isVerified, verificationErr := s.verifyMatch(ctx, client, baseUrl, serviceName, key)
+				s1.Verified = isVerified
+				if verificationErr != nil {
+					if errors.Is(verificationErr, noSuchHostErr) {
+						invalidHosts.Set(baseUrl, struct{}{})
+						continue EndpointLoop
+					}
+					s1.SetVerificationError(verificationErr, baseUrl)
+				}
+			}
+
+			results = append(results, s1)
+		}
+	}
+
+	return results, nil
+}
+
+func (s Scanner) Type() detectorspb.DetectorType {
+	return detectorspb.DetectorType_AzureDirectManagementKey
+}
+
+func (s Scanner) Description() string {
+	return "Azure API Management provides a direct management REST API for performing operations on selected entities, such as users, groups, products, and subscriptions."
+}
+
+func (s Scanner) IsFalsePositive(_ detectors.Result) (bool, string) {
+	return false, ""
+}
+
+func (s Scanner) verifyMatch(ctx context.Context, client *http.Client, baseUrl, serviceName, key string) (bool, error) {
+	url := fmt.Sprintf(
+		"%s/subscriptions/default/resourceGroups/default/providers/Microsoft.ApiManagement/service/%s/apis?api-version=2024-05-01",
+		baseUrl, serviceName,
+	)
+	accessToken, err := generateAccessToken(key)
+	if err != nil {
+		return false, err
+	}
+	req, err := http.NewRequestWithContext(ctx, "GET", url, nil)
+	if err != nil {
+		return false, err
+	}
+	req.Header.Set("Content-Type", "application/json")
+	req.Header.Set("Authorization", fmt.Sprintf("SharedAccessSignature %s", accessToken))
+	resp, err := client.Do(req)
+	if err != nil {
+		return false, nil
+	}
+	defer resp.Body.Close()
+
+	switch resp.StatusCode {
+	case http.StatusOK:
+		return true, nil
+	case http.StatusUnauthorized:
+		return false, nil
+	default:
+		return false, fmt.Errorf("unexpected HTTP response status %d", resp.StatusCode)
+	}
+}
+
+// https://learn.microsoft.com/en-us/rest/api/apimanagement/apimanagementrest/azure-api-management-rest-api-authentication
+func generateAccessToken(key string) (string, error) {
+	expiry := time.Now().UTC().Add(5 * time.Second).Format(RFC3339WithoutMicroseconds) // expire in 5 seconds
+	expiry = expiry + ".0000000Z"                                                      // 7 decimals microsecond's precision is must for access token
+
+	// Construct the string-to-sign
+	stringToSign := fmt.Sprintf("integration\n%s", expiry)
+
+	// Generate HMAC-SHA512 signature
+	h := hmac.New(sha512.New, []byte(key))
+	h.Write([]byte(stringToSign))
+	signature := h.Sum(nil)
+
+	// Base64 encode the signature
+	encodedSignature := base64.StdEncoding.EncodeToString(signature)
+
+	// Create the access token
+	accessToken := fmt.Sprintf("uid=integration&ex=%s&sn=%s", expiry, encodedSignature)
+	return accessToken, nil
+}

pkg/detectors/azuredirectmanagementkey/azuredirectmanagementkey_integration_test.go

@@ -0,0 +1,129 @@
+//go:build detectors
+// +build detectors
+
+package azuredirectmanagementkey
+
+import (
+	"context"
+	"fmt"
+	"testing"
+	"time"
+
+	"github.com/google/go-cmp/cmp"
+	"github.com/google/go-cmp/cmp/cmpopts"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/common"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/pb/detectorspb"
+)
+
+func TestAzureDirectManagementAPIKey_FromChunk(t *testing.T) {
+	ctx, cancel := context.WithTimeout(context.Background(), time.Second*5)
+	defer cancel()
+	testSecrets, err := common.GetSecret(ctx, "trufflehog-testing", "detectors5")
+	if err != nil {
+		t.Fatalf("could not get test secrets from GCP: %s", err)
+	}
+	url := testSecrets.MustGetField("AZUREDIRECTMANAGEMENTAPI_URL")
+	secret := testSecrets.MustGetField("AZUREDIRECTMANAGEMENTAPI_KEY")
+	inactiveSecret := testSecrets.MustGetField("AZUREDIRECTMANAGEMENTAPI_KEY_INACTIVE")
+
+	type args struct {
+		ctx    context.Context
+		data   []byte
+		verify bool
+	}
+	tests := []struct {
+		name                string
+		s                   Scanner
+		args                args
+		want                []detectors.Result
+		wantErr             bool
+		wantVerificationErr bool
+	}{
+		{
+			name: "found, verified",
+			s:    Scanner{},
+			args: args{
+				ctx:    ctx,
+				data:   []byte(fmt.Sprintf("You can find a azure management api url %s and key %s within", url, secret)),
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_AzureDirectManagementKey,
+					Verified:     true,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: false,
+		},
+		{
+			name: "found, unverified",
+			s:    Scanner{},
+			args: args{
+				ctx:    ctx,
+				data:   []byte(fmt.Sprintf("You can find a azure management api url %s and key %s within but not valid", url, inactiveSecret)), // the secret would satisfy the regex but not pass validation
+				verify: true,
+			},
+			want: []detectors.Result{
+				{
+					DetectorType: detectorspb.DetectorType_AzureDirectManagementKey,
+					Verified:     false,
+				},
+			},
+			wantErr:             false,
+			wantVerificationErr: false,
+		},
+		{
+			name: "not found",
+			s:    Scanner{},
+			args: args{
+				ctx:    ctx,
+				data:   []byte("You cannot find the secret within"),
+				verify: true,
+			},
+			want:                nil,
+			wantErr:             false,
+			wantVerificationErr: false,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got, err := tt.s.FromData(tt.args.ctx, tt.args.verify, tt.args.data)
+			if (err != nil) != tt.wantErr {
+				t.Errorf("AzureDirectManagementAPIKey.FromData() error = %v, wantErr %v", err, tt.wantErr)
+				return
+			}
+			for i := range got {
+				if len(got[i].Raw) == 0 {
+					t.Fatalf("no raw secret present: \n %+v", got[i])
+				}
+				if (got[i].VerificationError() != nil) != tt.wantVerificationErr {
+					t.Fatalf("wantVerificationError = %v, verification error = %v", tt.wantVerificationErr, got[i].VerificationError())
+				}
+			}
+			ignoreOpts := cmpopts.IgnoreFields(detectors.Result{}, "Raw", "RawV2", "Redacted", "verificationError")
+			if diff := cmp.Diff(got, tt.want, ignoreOpts); diff != "" {
+				t.Errorf("AzureDirectManagementAPIKey.FromData() %s diff: (-got +want)\n%s", tt.name, diff)
+			}
+		})
+	}
+}
+
+func BenchmarkFromData(benchmark *testing.B) {
+	ctx := context.Background()
+	s := Scanner{}
+	for name, data := range detectors.MustGetBenchmarkData() {
+		benchmark.Run(name, func(b *testing.B) {
+			b.ResetTimer()
+			for n := 0; n < b.N; n++ {
+				_, err := s.FromData(ctx, false, data)
+				if err != nil {
+					b.Fatal(err)
+				}
+			}
+		})
+	}
+}

pkg/detectors/azuredirectmanagementkey/azuredirectmanagementkey_test.go

@@ -0,0 +1,86 @@
+package azuredirectmanagementkey
+
+import (
+	"context"
+	"testing"
+
+	"github.com/google/go-cmp/cmp"
+
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/engine/ahocorasick"
+)
+
+var (
+	validPattern = `
+	AZURE_MANGEMENT_API_KEY=UJh1Wn7txjls2GPK1YxO9+3tpqQffSfxb+97PmT8j3cSQoXvGa74lCKpBqPeppTHCharbaMeKqKs/H4gA/go1w==
+	AZURE_MANAGEMENT_API_URL=https://trufflesecuritytest.management.azure-api.net
+	`
+	invalidPattern = `
+	AZURE_MANGEMENT_API_KEY=UJh1Wn7txjls2GPK1YxO9+3tpqQffSfxb+97PmT8j3cSQoXvGa74lCKp
+	AZURE_MANAGEMENT_API_URL=https://trufflesecuritytest.management.azure-api.net
+	`
+)
+
+func TestAzureDirectManagementAPIKey_Pattern(t *testing.T) {
+	d := Scanner{}
+	ahoCorasickCore := ahocorasick.NewAhoCorasickCore([]detectors.Detector{d})
+
+	tests := []struct {
+		name  string
+		input string
+		want  []string
+	}{
+		{
+			name:  "valid pattern",
+			input: validPattern,
+			want:  []string{"https://trufflesecuritytest.management.azure-api.net:UJh1Wn7txjls2GPK1YxO9+3tpqQffSfxb+97PmT8j3cSQoXvGa74lCKpBqPeppTHCharbaMeKqKs/H4gA/go1w=="},
+		},
+		{
+			name:  "invalid pattern",
+			input: invalidPattern,
+			want:  nil,
+		},
+	}
+
+	for _, test := range tests {
+		t.Run(test.name, func(t *testing.T) {
+			matchedDetectors := ahoCorasickCore.FindDetectorMatches([]byte(test.input))
+			if len(matchedDetectors) == 0 {
+				t.Errorf("keywords '%v' not matched by: %s", d.Keywords(), test.input)
+				return
+			}
+
+			results, err := d.FromData(context.Background(), false, []byte(test.input))
+			if err != nil {
+				t.Errorf("error = %v", err)
+				return
+			}
+
+			if len(results) != len(test.want) {
+				if len(results) == 0 {
+					t.Errorf("did not receive result")
+				} else {
+					t.Errorf("expected %d results, only received %d", len(test.want), len(results))
+				}
+				return
+			}
+
+			actual := make(map[string]struct{}, len(results))
+			for _, r := range results {
+				if len(r.RawV2) > 0 {
+					actual[string(r.RawV2)] = struct{}{}
+				} else {
+					actual[string(r.Raw)] = struct{}{}
+				}
+			}
+			expected := make(map[string]struct{}, len(test.want))
+			for _, v := range test.want {
+				expected[v] = struct{}{}
+			}
+
+			if diff := cmp.Diff(expected, actual); diff != "" {
+				t.Errorf("%s diff: (-want +got)\n%s", test.name, diff)
+			}
+		})
+	}
+}

pkg/engine/defaults/defaults.go

@@ -77,6 +77,7 @@ import (
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/azureapimanagementsubscriptionkey"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/azurecontainerregistry"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/azuredevopspersonalaccesstoken"
+	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/azuredirectmanagementkey"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/azuresastoken"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/azuresearchadminkey"
 	"github.com/trufflesecurity/trufflehog/v3/pkg/detectors/azuresearchquerykey"
@@ -916,6 +917,7 @@ func buildDetectorList() []detectors.Detector {
 		&azure_cosmosdb.Scanner{},
 		&azurecontainerregistry.Scanner{},
 		&azuredevopspersonalaccesstoken.Scanner{},
+		&azuredirectmanagementkey.Scanner{},
 		// &azurefunctionkey.Scanner{}, // detector is throwing some FPs
 		&azure_openai.Scanner{},
 		&azuresastoken.Scanner{},

pkg/pb/detectorspb/detectors.pb.go

@@ -1128,6 +1128,7 @@ const (
 	DetectorType_Langfuse                                DetectorType = 1021
 	DetectorType_BingSubscriptionKey                     DetectorType = 1022
 	DetectorType_XAI                                     DetectorType = 1023
+	DetectorType_AzureDirectManagementKey                DetectorType = 1024
 )
 
 // Enum value maps for DetectorType.
@@ -2153,6 +2154,7 @@ var (
 		1021: "Langfuse",
 		1022: "BingSubscriptionKey",
 		1023: "XAI",
+		1024: "AzureDirectManagementKey",
 	}
 	DetectorType_value = map[string]int32{
 		"Alibaba":                               0,
@@ -3175,6 +3177,7 @@ var (
 		"Langfuse":                          1021,
 		"BingSubscriptionKey":               1022,
 		"XAI":                               1023,
+		"AzureDirectManagementKey":          1024,
 	}
 )
 
@@ -3628,7 +3631,7 @@ var file_detectors_proto_rawDesc = []byte{
 	0x4c, 0x41, 0x49, 0x4e, 0x10, 0x01, 0x12, 0x0a, 0x0a, 0x06, 0x42, 0x41, 0x53, 0x45, 0x36, 0x34,
 	0x10, 0x02, 0x12, 0x09, 0x0a, 0x05, 0x55, 0x54, 0x46, 0x31, 0x36, 0x10, 0x03, 0x12, 0x13, 0x0a,
 	0x0f, 0x45, 0x53, 0x43, 0x41, 0x50, 0x45, 0x44, 0x5f, 0x55, 0x4e, 0x49, 0x43, 0x4f, 0x44, 0x45,
-	0x10, 0x04, 0x2a, 0xdc, 0x83, 0x01, 0x0a, 0x0c, 0x44, 0x65, 0x74, 0x65, 0x63, 0x74, 0x6f, 0x72,
+	0x10, 0x04, 0x2a, 0xfb, 0x83, 0x01, 0x0a, 0x0c, 0x44, 0x65, 0x74, 0x65, 0x63, 0x74, 0x6f, 0x72,
 	0x54, 0x79, 0x70, 0x65, 0x12, 0x0b, 0x0a, 0x07, 0x41, 0x6c, 0x69, 0x62, 0x61, 0x62, 0x61, 0x10,
 	0x00, 0x12, 0x08, 0x0a, 0x04, 0x41, 0x4d, 0x51, 0x50, 0x10, 0x01, 0x12, 0x07, 0x0a, 0x03, 0x41,
 	0x57, 0x53, 0x10, 0x02, 0x12, 0x09, 0x0a, 0x05, 0x41, 0x7a, 0x75, 0x72, 0x65, 0x10, 0x03, 0x12,
@@ -4682,11 +4685,13 @@ var file_detectors_proto_rawDesc = []byte{
 	0x0d, 0x0a, 0x08, 0x4c, 0x61, 0x6e, 0x67, 0x66, 0x75, 0x73, 0x65, 0x10, 0xfd, 0x07, 0x12, 0x18,
 	0x0a, 0x13, 0x42, 0x69, 0x6e, 0x67, 0x53, 0x75, 0x62, 0x73, 0x63, 0x72, 0x69, 0x70, 0x74, 0x69,
 	0x6f, 0x6e, 0x4b, 0x65, 0x79, 0x10, 0xfe, 0x07, 0x12, 0x08, 0x0a, 0x03, 0x58, 0x41, 0x49, 0x10,
-	0xff, 0x07, 0x42, 0x3d, 0x5a, 0x3b, 0x67, 0x69, 0x74, 0x68, 0x75, 0x62, 0x2e, 0x63, 0x6f, 0x6d,
-	0x2f, 0x74, 0x72, 0x75, 0x66, 0x66, 0x6c, 0x65, 0x73, 0x65, 0x63, 0x75, 0x72, 0x69, 0x74, 0x79,
-	0x2f, 0x74, 0x72, 0x75, 0x66, 0x66, 0x6c, 0x65, 0x68, 0x6f, 0x67, 0x2f, 0x76, 0x33, 0x2f, 0x70,
-	0x6b, 0x67, 0x2f, 0x70, 0x62, 0x2f, 0x64, 0x65, 0x74, 0x65, 0x63, 0x74, 0x6f, 0x72, 0x73, 0x70,
-	0x62, 0x62, 0x06, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x33,
+	0xff, 0x07, 0x12, 0x1d, 0x0a, 0x18, 0x41, 0x7a, 0x75, 0x72, 0x65, 0x44, 0x69, 0x72, 0x65, 0x63,
+	0x74, 0x4d, 0x61, 0x6e, 0x61, 0x67, 0x65, 0x6d, 0x65, 0x6e, 0x74, 0x4b, 0x65, 0x79, 0x10, 0x80,
+	0x08, 0x42, 0x3d, 0x5a, 0x3b, 0x67, 0x69, 0x74, 0x68, 0x75, 0x62, 0x2e, 0x63, 0x6f, 0x6d, 0x2f,
+	0x74, 0x72, 0x75, 0x66, 0x66, 0x6c, 0x65, 0x73, 0x65, 0x63, 0x75, 0x72, 0x69, 0x74, 0x79, 0x2f,
+	0x74, 0x72, 0x75, 0x66, 0x66, 0x6c, 0x65, 0x68, 0x6f, 0x67, 0x2f, 0x76, 0x33, 0x2f, 0x70, 0x6b,
+	0x67, 0x2f, 0x70, 0x62, 0x2f, 0x64, 0x65, 0x74, 0x65, 0x63, 0x74, 0x6f, 0x72, 0x73, 0x70, 0x62,
+	0x62, 0x06, 0x70, 0x72, 0x6f, 0x74, 0x6f, 0x33,
 }
 
 var (

proto/detectors.proto

@@ -1033,6 +1033,7 @@ enum DetectorType {
   Langfuse = 1021;
   BingSubscriptionKey = 1022;
   XAI = 1023;
+  AzureDirectManagementKey = 1024;
 }
 
 message Result {